MCPcopy Create free account

hub / github.com/YutoTerashima/agent-safety-eval-lab / functions

Functions120 in github.com/YutoTerashima/agent-safety-eval-lab

↓ 20 callersFunctionsave_bar
(path: Path, labels: list[str], values: list[float], title: str, ylabel: str)
scripts/run_matrix.py:150
↓ 16 callersFunctionwrite_json
(path: Path, payload: Any)
scripts/run_matrix.py:78
↓ 12 callersFunctionwrite_csv
(path: Path, rows: list[dict[str, Any]])
scripts/run_matrix.py:83
↓ 10 callersFunctionensure_dirs
()
scripts/gpu_experiment.py:38
↓ 9 callersFunctionnormalize_text
(text: str)
scripts/run_matrix.py:100
↓ 8 callersFunctionload_config
()
scripts/gpu_experiment.py:34
↓ 8 callersFunctionpreprocess_data
(max_samples: int | None = None)
scripts/gpu_experiment.py:139
↓ 7 callersFunctionbuild_index
(cfg: dict[str, Any], results: list[dict[str, Any]], artifacts: list[str], profile: str, device: str)
scripts/run_matrix.py:670
↓ 7 callersFunctiondevice_info
(requested: str)
scripts/gpu_experiment.py:43
↓ 7 callersFunctionread_jsonl
(path: Path)
scripts/gpu_experiment.py:117
↓ 5 callersFunctionnormalize_text
(value: Any)
scripts/gpu_experiment.py:59
↓ 5 callersFunctionread_jsonl
(path: Path, limit: int | None = None)
scripts/run_matrix.py:67
↓ 5 callersFunctionwrite_jsonl
(path: Path, rows: list[dict[str, Any]])
scripts/gpu_experiment.py:112
↓ 4 callersFunctionredact_for_release
(text: str, max_chars: int = 260)
scripts/run_matrix.py:104
↓ 3 callersFunctionclassification_metrics
(y_true, y_pred, labels: list[str], score=None)
scripts/run_matrix.py:209
↓ 3 callersFunctiondownload_data
(max_samples: int, smoke: bool)
scripts/gpu_experiment.py:121
↓ 3 callersFunctionfailure_text_for_release
(text: str, true_label: str | int, pred_label: str | int)
scripts/run_matrix.py:117
↓ 3 callersFunctionrecall_for
(y_true, y_pred, label: str)
scripts/run_matrix.py:361
↓ 3 callersFunctionrule_predict
(texts: list[str])
scripts/run_matrix.py:200
↓ 3 callersFunctionrun_demo
(dataset_path: Path | None = None)
src/agent_safety_eval_lab/mock_runner.py:43
↓ 3 callersFunctiontokenize
(text: str)
scripts/gpu_experiment.py:291
↓ 2 callersFunction_as_float
(value: Any, default: float = 0.0)
src/agent_safety_eval_lab/calibration.py:19
↓ 2 callersFunction_as_float
(value: Any, default: float = 0.0)
src/agent_safety_eval_lab/safety_classifier.py:19
↓ 2 callersFunction_as_float
(value: Any, default: float = 0.0)
src/agent_safety_eval_lab/trace_policy.py:19
↓ 2 callersFunction_metric
(row: Mapping[str, Any], *names: str)
src/agent_safety_eval_lab/calibration.py:26
↓ 2 callersFunction_metric
(row: Mapping[str, Any], *names: str)
src/agent_safety_eval_lab/safety_classifier.py:26
↓ 2 callersFunction_metric
(row: Mapping[str, Any], *names: str)
src/agent_safety_eval_lab/trace_policy.py:26
↓ 2 callersFunctionbuild_markdown
(summary: dict[str, int], lines: list[str])
src/agent_safety_eval_lab/report.py:9
↓ 2 callersFunctionencode_texts
(texts: list[str], vocab: dict[str, int], seq_len: int)
scripts/run_matrix.py:545
↓ 2 callersFunctiongrade_trace
(trace: AgentTrace, case: EvalCase)
src/agent_safety_eval_lab/grader.py:6
↓ 2 callersFunctionload_cases
(path: Path)
src/agent_safety_eval_lab/mock_runner.py:10
↓ 2 callersFunctionload_classification
(cfg: dict[str, Any], profile: str)
scripts/run_matrix.py:174
↓ 2 callersFunctionmain
()
scripts/gpu_experiment.py:619
↓ 2 callersFunctionmock_agent
(case: EvalCase)
src/agent_safety_eval_lab/mock_runner.py:23
↓ 2 callersFunctionperturb_text
(text: str, name: str)
scripts/run_matrix.py:259
↓ 2 callersFunctionread_json
(path: Path, default)
scripts/make_report.py:10
↓ 2 callersFunctionreplay_directory
(trace_dir: Path)
src/agent_safety_eval_lab/replay.py:46
↓ 2 callersFunctionrows_to_examples
(rows: list[dict[str, Any]], cfg: dict[str, Any])
scripts/gpu_experiment.py:75
↓ 2 callersFunctionrun_experiment
(dataset_path: Path | None = None)
src/agent_safety_eval_lab/experiment.py:11
↓ 2 callersFunctionrun_gpu_classifier
(examples: list[dict[str, Any]], device: torch.device, epochs: int, seed: int = 7)
scripts/gpu_experiment.py:196
↓ 2 callersFunctionsave_heatmap
(path: Path, matrix: np.ndarray, xlabels: list[str], ylabels: list[str], title: str)
scripts/run_matrix.py:162
↓ 2 callersFunctionsuite_coverage
(cases: list[dict])
src/agent_safety_eval_lab/coverage.py:9
↓ 2 callersFunctiontoken_set
(text: str)
scripts/run_matrix.py:125
↓ 2 callersFunctionwrite_experiment_outputs
(report_dir: Path | None = None)
src/agent_safety_eval_lab/experiment.py:39
↓ 1 callersFunctionaggregate_results
(results: list[EvalResult])
src/agent_safety_eval_lab/analytics.py:44
↓ 1 callersFunctionanalyze_trace
(trace: AgentTrace, result: EvalResult)
src/agent_safety_eval_lab/analytics.py:28
↓ 1 callersFunctionbuild_vocab
(texts: list[str], max_vocab: int)
scripts/run_matrix.py:535
↓ 1 callersMethoddecide
(self, tool: str, arguments: dict[str, str], prior_calls: int = 0)
src/agent_safety_eval_lab/policy.py:27
↓ 1 callersFunctiondevice_info
(requested: str)
scripts/run_matrix.py:47
↓ 1 callersFunctionensure_dirs
()
scripts/run_matrix.py:62
↓ 1 callersFunctionfetch_dataset
(cfg: dict[str, Any], max_samples: int, smoke: bool)
scripts/gpu_experiment.py:87
↓ 1 callersFunctiongit_commit
()
scripts/run_matrix.py:40
↓ 1 callersFunctionlabel_from_row
(row: dict[str, Any], expr: str)
scripts/gpu_experiment.py:69
↓ 1 callersFunctionlcs_len
(a: list[str], b: list[str])
scripts/run_matrix.py:129
↓ 1 callersFunctionload_trace
(path: Path)
src/agent_safety_eval_lab/replay.py:10
↓ 1 callersFunctionmain
()
src/agent_safety_eval_lab/cli.py:11
↓ 1 callersFunctionmain
()
scripts/run_matrix.py:685
↓ 1 callersFunctionmain
()
scripts/make_report.py:14
↓ 1 callersFunctionmain
()
scripts/analyze_failures.py:11
↓ 1 callersFunctionmake_report
()
scripts/gpu_experiment.py:558
↓ 1 callersFunctionndcg_at
(rank: int)
scripts/gpu_experiment.py:295
↓ 1 callersFunctionrank_classifiers
Return rows ranked by the strongest available quality signal.
src/agent_safety_eval_lab/safety_classifier.py:33
↓ 1 callersFunctionread_matrix
(path: Path)
scripts/run_matrix.py:36
↓ 1 callersFunctionrender_dashboard
(traces: list[dict[str, Any]], title: str)
scripts/run_matrix.py:494
↓ 1 callersFunctionreplay_trace
(trace: dict)
src/agent_safety_eval_lab/replay.py:14
↓ 1 callersFunctionrouge_l
(pred: str, ref: str)
scripts/run_matrix.py:140
↓ 1 callersFunctionroute_trace
Return rows ranked by the strongest available quality signal.
src/agent_safety_eval_lab/trace_policy.py:33
↓ 1 callersFunctionrun_cards
(args: argparse.Namespace)
scripts/gpu_experiment.py:528
↓ 1 callersFunctionrun_cards
(cfg: dict[str, Any], profile: str, device: str)
scripts/run_matrix.py:628
↓ 1 callersFunctionrun_classification
(cfg: dict[str, Any], profile: str, device: str)
scripts/run_matrix.py:279
↓ 1 callersFunctionrun_classification_experiment
(args: argparse.Namespace)
scripts/gpu_experiment.py:269
↓ 1 callersFunctionrun_eval_methods
(args: argparse.Namespace)
scripts/gpu_experiment.py:357
↓ 1 callersFunctionrun_eval_methods
(cfg: dict[str, Any], profile: str, device: str)
scripts/run_matrix.py:440
↓ 1 callersFunctionrun_prompt_robustness
(cfg: dict[str, Any], profile: str, device: str)
scripts/run_matrix.py:597
↓ 1 callersFunctionrun_rag
(cfg: dict[str, Any], profile: str, device: str)
scripts/run_matrix.py:369
↓ 1 callersFunctionrun_retrieval_experiment
(args: argparse.Namespace)
scripts/gpu_experiment.py:299
↓ 1 callersFunctionrun_robustness
(args: argparse.Namespace)
scripts/gpu_experiment.py:462
↓ 1 callersFunctionrun_tiny_transformer
(args: argparse.Namespace)
scripts/gpu_experiment.py:389
↓ 1 callersFunctionrun_trace
(args: argparse.Namespace)
scripts/gpu_experiment.py:494
↓ 1 callersFunctionrun_trace
(cfg: dict[str, Any], profile: str, device: str)
scripts/run_matrix.py:465
↓ 1 callersFunctionrun_transformer
(cfg: dict[str, Any], profile: str, device: str)
scripts/run_matrix.py:499
↓ 1 callersFunctionsave_confusion
(labels: list[str], y_true: list[str], y_pred: list[str], path: Path)
scripts/gpu_experiment.py:178
↓ 1 callersFunctionstable_hash
(text: str)
scripts/run_matrix.py:95
↓ 1 callersFunctionsummarize
(results: list[EvalResult])
src/agent_safety_eval_lab/grader.py:38
↓ 1 callersFunctionsynthetic_predictions
(row: dict[str, Any])
scripts/run_matrix.py:427
↓ 1 callersFunctionthreshold_sweep
Return rows ranked by the strongest available quality signal.
src/agent_safety_eval_lab/calibration.py:33
↓ 1 callersFunctiontorch_mlp_train
(x_train, y_train, x_test, labels: list[str], device: str, epochs: int = 5)
scripts/run_matrix.py:228
↓ 1 callersFunctiontrain_tiny_transformer
(train_texts, train_y, test_texts, vocab, exp, device, epochs: int)
scripts/run_matrix.py:568
↓ 1 callersFunctionwrite_report
(report_dir: Path | None = None)
src/agent_safety_eval_lab/report.py:43
Method__init__
(self, rules: list[PolicyRule])
src/agent_safety_eval_lab/policy.py:24
Method__init__
(self, vocab_size: int)
scripts/gpu_experiment.py:417
Method__init__
(self, vocab_size: int, dim: int, layers: int, heads: int, dropout: float)
scripts/run_matrix.py:554
Functionbest_unsafe_recall
Summarize the project-specific tradeoff metrics used in the report.
src/agent_safety_eval_lab/safety_classifier.py:44
Functioncalibration_gap
Create a compact count table for errors, routes, groups, or artifacts.
src/agent_safety_eval_lab/safety_classifier.py:59
Methodforward
(self, ids)
scripts/gpu_experiment.py:423
Methodforward
(self, x)
scripts/run_matrix.py:560
Functionmaturity_claims
()
src/agent_safety_eval_lab/research_report.py:28
Functionoperating_points
Create a compact count table for errors, routes, groups, or artifacts.
src/agent_safety_eval_lab/calibration.py:59
Functionpytest_configure
(config)
tests/conftest.py:5
Functionrecommended_threshold
Summarize the project-specific tradeoff metrics used in the report.
src/agent_safety_eval_lab/calibration.py:44
next →1–100 of 120, ranked by callers